Duskblade, AD Assassins, and You

Hi folks,

We’ve put a new item on the PBE aimed at solving a few of the problems AD-scaling Assassins face when interacting with the item system.

The current item details are:

Duskblade of Draktharr

• Recipe: Serrated Dirk + B.F. Sword + 850 gold (3250 total gold)
• +75 Attack Damage
• +5% Movement Speed
• UNIQUE Passive: +10 Armor Penetration
• UNIQUE Passive: Basic attacks on an enemy champion apply Nightfall (120 second cooldown).
• Nightfall: After 2 seconds, deal physical damage equal to 90 plus 25% of the target's missing health. If you get a kill or assist on the target before Nightfall ends, the cooldown is refunded.

I put a preliminary post on this up in December here when we were a bit earlier in iteration: http://boards.na.leagueoflegends.com/en/c/developer-corner/GhWdm3R6-new-ad-assassin-item-in-internal-testing

First, the obvious. AD assassins aren’t, as a group, weak at this point. A few aren’t performing well, and a few are performing a bit too well. That said, the challenge we’re trying to solve with Duskblade isn’t one of strength, even though this item will definitely impact character strength.

This will make more sense in a chain of events as they currently happen in League (I’ll explain the points in a sec):

• There is poor itemization support for AD assassins (2-4 core items + random AD items)
• An AD assassin champion is made and doesn’t succeed because of the above
• We build additional power into the AD assassin’s kit
• AD Assassins that succeed are the ones with more levels or synergize well with the current hodgepodge of AD items

… all of the above leads us to our current approach, which is:

• By building real AD assassin items, we can actually gauge an AD assassin’s performance along the same item progression as other champions

… which means that we can balance for those champions who are too weak or strong.

Breaking it down:

AD Assassins don’t feel well-supported by the item system...

Once again, not champion strength. I’m talking about the feeling of support. We’ve talked a lot about choice in the item system lately (e.g. the Marksman item rework), but that’s only one of the critical values provided by our item system. Another is adequately rewarding successes - something that our item system does pretty poorly for AD assassin champions.

AD assassin builds typically are made of 2-4 items that suit them very well (exactly which items depends on the champion) and then a grab bag of whatever else they can find. The net result is that they’re rewarded a LOT for early success, but aren’t rewarded nearly as much while they’re slogging through their last couple items.

... As a result we overcompensate on their kits.

The other side of the above problem is that we have to balance these champions’ power levels around items that actually exist.

Champions get meaningfully less power from items that are not optimized for them - an ADC gets less value out of Dead Man’s Plate than Tanks and Fighters do, and less out of Zhonya’s Hourglass than Mages. Because ADCs are fully supported with their own items, this is never a problem for them. If there are a few fringe cases where Dead Man’s Plate or Zhonya’s Hourglass is actually a good idea, it’s because they’re the optimal tool for very specific task, but the tradeoff needs to be significant (ie: loss of a 6th item slot for essentially cross-role ‘splash’ utility).

AD Assassins aren’t similarly supported. They have too few items that are optimized for them, which means that they effectively just get less power - in many cases, much less power - out of the gold they spend later in a game. In order to balance them, that difference in power has to be compensated somewhere.

We have a few main ways of doing that. Some AD assassins have fairly strong (or safe) lane phases that let them maintain a gold lead. So they might be getting less value out of gold (due to poor itemization), but this is offset by the fact that they have more gold to spend. This means, in turn, we get a lot of snowball champions. This isn’t inherently bad for the game, but not something we should be forced to do for all champions of a given class.

Final point: if a champion’s kit can scale better later in the game than usual, we need to be aware of the tradeoff we’re making because it’s much more difficult to deny an opponent experience than gold. As a result, their power spike later in the game becomes extremely reliable (compared to a champion who relies on items).

By fully supporting AD assassins in the item system, we no longer have to balance them around the assumption that their kits will be making up a sizable power deficit from items, which means that their kits can have either less guaranteed early games or less guaranteed late games, depending on the champion.

And there’s collateral damage for item changes

Because AD Assassins are cobbling together items from other classes, they’re very sensitive to changes done for other classes. For example, the Last Whisper changes at preseason hurt them quite badly even though they were generally successful for other classes of users - an Assassin’s best targets have very little bonus armor.

By supporting AD assassins with distinct itemization, we both minimize cases where we accidentally damage their state (due to work on other parts of the game) and increase our ability to tune them when they’re strong or weak as a group.

We can be up front about this, however: Duskblade is intended to be a core item for a lot of assassins (a first, second, or third buy, depending) and we will be balancing around it. That said, we do this anyway based on other items, it’s just a lot more random which items we balance which AD assassins around.

... But AD assassins are frustrating.

We understand that AD assassins offer low counterplay and significant frustration playing against and, by working on them, we risk exacerbating that in the short term. We feel this work is important because it puts us in a better position to address those pain points as well.